Wikidata:Weekend Workshop for Wikipedians

This report documents a workshop in Cologne (de:Wikipedia:Lokal_K/Wikidata-Workshop), that was held in May 2018. It can be used as a template for similar workshops.

Participants and Instructors

12 participants, all experienced editors in Wikipedia and Wikimedia Commons, but with varying degrees of experience with Wikidata. While some participants had never edited in Wikidata, most participants had done at least some small scale editing before the workshop. The event was presented by Lucas Werkmeister and Matti Blume with organizational support from WMDE's Community Support Team.

Preparation

Preparation was supported by Nichole Saad from the Wikipedia Education Program, to achieve better learning outcome through a backwards planning approach. For this we first specified the learning goals, then identified the evidence, we need to know that those goals were met and lastly planned the activities and timetable around that.

The specified learning goals for participants were:

Understanding and adding Data

Recall the basic structure of items, properties and statements on Wikidata

Explain how items on Wikidata are interlinked

Demonstrate understanding of items

Add data in bulk using QuickStatements

Using data

Interpret data they find on Wikidata

Understand how to write a query

Use a query to generate a list

Produce a query

Those learning goals would be verified in one of two ways:

1. Upload a batch of data in bulk based on their interests and check it using the query.
   • structure of the data is sound
   • correct information is uploaded
2. Automatically create a list on Wikipedia based on a query in their userspace.

Furthermore we asked participants beforehand about their expectations and needs. Most stated specific projects or topics, where they would like make better use of Wikidata.

To evaluate the outcome of the workshop a brief exit survey (7 questions, all optional) was prepared, using Google Forms (see #Results).

Resources

• 2 presenters (very experienced in query service and QuickStatements)
• video projector
• attendants BYOD + loaner laptops for participants without own device
• some printed copies of Wikidata in Brief and Wikidata Query Service in Brief

Timetable

Day 1

based on Wikidata:Planning_a_Wikidata_workshop#Structure_of_a_Workshop

Morning: Introduction

• What is Wikidata (why is it cool)?
  • specific examples:
    • Histropedia Empires
    • fr.wikinews (Olympia data) – item
    • Reasonator

• Statements – decent example item: Wikipedia:Lokal K (Q17285924)
  • Explain the property (instance of (P31))
  • Explain the value
  • Values can be many things

• Activity: look up something (sport, work of art, street names), share all the statements and values you find there with a partner

• Basic editing: explain adding a basic statement (property + value) to an existing item

• Activity: find an item of some human with a missing given or family name and add that statement

• “What links here”: show Special:WhatLinksHere (from sidebar) from a name item to show the links between items, then show in which statement they are linked

• Explain and demonstrate qualifiers in item Cologne (Q365):
  • name examples of “important” qualifiers (start time (P580), end time (P582), point in time (P585), series ordinal (P1545));
  • technically every property can be used as qualifier

• Explain and demonstrate references
  • reference URL (P854), stated in (P248), retrieved (P813), etc.

• Activity: add a statement with a reference to an item

• Long term activity: explore your personal area of interest using the the skill you just learned

Afternoon: Query Service

Start with some motivational examples, based on the interests of the participants which they had previously announced on the workshop page:

• streets in Berlin with gender of the person they’re named after
• paintings by Lovis Corinth
• taxonomy graph of Pantherinae
• schools, universities and kindergartens around the Lokal K

Introduce the query helper: find paintings by Georges Kars.

Show different result views.

The plan was to mostly stick to the query helper, and ideally not use SPARQL at all, since that would be beyond the scope of the workshop. However, it quickly became clear that the query helper was too limited for the needs of the participants, and we started exploring SPARQL. Some of the participants had already worked with SPARQL a bit, and were able to explain some aspects in a different, more accessible way.

In some intervals, participants wrote their own queries, and presenters walked around and helped out where problems arose.

Some of the resulting queries can be found on the workshop page.

We also showed how to generate on-wiki lists from Wikidata using {{Wikidata list}} (aka Listeria), and demonstrated it on the workshop page.

Day 2

Morning: Data uploads

• ask audience about their data, they want to upload
• introduction to data sources:
• Bring your own data
  • briefly explain different formats (XLS, CSV, etc) and their use
• present some open data portals (https://daten.berlin.de/, https://data.europa.eu/euodp/en/home/, https://www.govdata.de/)
• Introduction to PetScan
  • example of data collection from Wikipedia through PetScan based on some of the participants' topics
  • activity: write a small batch to Wikidata with Petscan
• QuickStatements 2:
  • brief introduction to data cleaning with Excel/Calc/gSheets
    • especially VLOOKUP
    • explain the two possible input formats for QuickStatements and present one example batch for each
• Check uploads with queries for completeness, errors etc.

Afternoon: free Wikidata editing

Both presenters are available for everyone, who wants to edit on their own. Can answer question, help with queries.

Results and suggestions for similar events

The exit survey asked participants, whether they have edited Wikidata before the workshop, if their expectations were met, whether they plan to keep editing Wikidata and to name the best and worst aspects about the workshop.

• All participants, who answered the survey stated, that their expectations about the workshop were at least met, while most stated they were somewhat surpassed.
• Almost all participants plan to keep editing and/or using data in Wikidata after the workshop.
• All participants were especially interested in the use of the query service, while most were also especially interested in mass editing with tools.
• The learner centric approach of the workshop was complimented
• Wishes were made towards a follow-up workshops a few months later, to deep dive into some of the tools, that were presented at this workshop.

Another learning of this workshop was, that the more the participants are engaged in the activities, the longer each step took, because of questions. Therefore, it is advised to plan with at least 50 % more time, than what intuition would say, how long an activity would take.

According to the survey, Wikidata merchandise actually helps to increase the motivation levels. Bring branded materials (T-shirts, notepads, lapel pins, stickers, etc.) and offer it to participants if possible.